The healthcare burden of opioid abuse is substantial; abusers often have complex healthcare needs and may require care beyond that which is required to treat abuse.
Objectives: To replicate and extend a recently published analysis of the drivers of excess costs of opioid abuse.
Study Design: Retrospective data analysis using de-identified claims data from the Truven MarketScan Commercial Claims and Encounter database.
Methods: Medical and prescription drug claims from beneficiaries covered by large self-insured US companies were used to select patients with incident diagnoses of opioid abuse between 2012 and 2015. Two cohorts, abusers and nonabusers, were matched using propensity score methods. Excess healthcare costs were estimated over a 6-month baseline period and 12-month follow-up period. Cost drivers were assessed by diagnosis (3-digit International Classification of Diseases, Ninth Revision, Clinical Modification groupings) and place of service.
Results: The analysis included 73,714 matched pairs of abusers and nonabusers. Relative to nonabusers, abusers had considerably higher annual healthcare costs of $10,989 per patient, or $1.98 per member per month. Excess costs were similar, yet lower, than the previous analysis using another commercial claims database. In both analyses, a ramp-up in excess costs was observed prior to the incident abuse diagnosis, followed by a decline post diagnosis, although not to baseline levels. Key drivers of excess costs in the 2 studies included opioid use disorders, nonopioid substance misuse, and painful and mental health conditions. From 2010 to 2014, the prevalence of diagnosed opioid abuse doubled, with incidence rates exhibiting an increasing, though flatter, trend than earlier in the period.
Conclusions: Opioid abuse imposes a considerable economic burden on payers. Many abusers have complex healthcare needs and may require care beyond that which is required to treat opioid abuse. These results are robust and consistent across different data sources.
Prescription opioid pain relievers can be highly effective in providing relief for patients suffering from chronic pain.1 At the same time, prescription opioid abuse, dependence, overdose, and poisoning (hereinafter “abuse”) have collectively become a national public health concern. Policy interventions, such as states’ prescription drug monitoring programs and rescheduling certain opioids, as well as technological innovations, such as the introduction of abuse-deterrent formulations, have been implemented with the goal of curbing opioid misuse while trying to maintain appropriate access to care for patients.
Opioid abuse is also costly; societal costs of opioid abuse in 2007 were estimated at $55.7 billion.2 Payers face a large proportion of these costs; previous estimates of the annual per-patient economic burden to payers of opioid abuse range from $10,000 to $20,000.3-5 Despite growing concern, the current literature has only begun to address the drivers of these excess costs of opioid abuse in terms of place of service and comorbid conditions.
This paper first seeks to replicate the findings of a newly published study performed using a commercial claims database from OptumHealth Care Solutions, Inc (Optum). The original study presented recent refined estimates of the excess costs of opioid abuse from a payer perspective and investigated the trajectory and drivers of those excess costs. The goal of this study was to replicate the findings using a different, more expansive database: the Truven MarketScan Commercial Claims and Encounter database (Truven). In addition, we extend those findings by presenting updated rates of diagnosed prevalence and incidence of opioid abuse, as well as the incremental per-member-per-month (PMPM) cost of abuse to payers.
We analyzed de-identified administrative claims data from Truven, which cover more than one-third of the US population (ie, 117 million commercially insured beneficiaries, including employees, children, spouses, and retirees) of large, self-insured companies located throughout the United States and in a broad array of industries and job types. The databases aggregate data from a large number of plan types (eg, fee-for-service and health maintenance organizations [HMOs]).6 The data include medical claims (service date, diagnoses received, procedures performed, place of service, and payment amounts), pharmacy claims (fill dates, national drug codes, and payment amounts), and eligibility information (patient demographic characteristics and enrollment history) for the period of January 2010 to September 2015. The Optum data used in the prior study contained a similar commercially insured population from different companies, industries, and regions.
Study Design and Statistical Analysis
To ensure replicability, the methodology applied for the replication analysis paralleled that in the original analysis to the extent possible. Full details of the sample selection procedure and baseline characteristics can be found in the original study.7
Incident abuse diagnoses between January 2012 and September 2015 were identified in the medical claims using the following International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for opioid abuse, dependence, or overdose/poisoning: 304.0x, 304.7x, 305.5x, 965.00, 965.02, and 965.09.7 The index date for abusers was defined as the date of the first abuse, dependence, or overdose/poisoning diagnosis. For nonabusers, the index date was the date of a random medical claim during this same period.
The follow-up period over which each individual was observed was defined as the 6-month period prior to the index date and 0 to 6 months thereafter. Including the 6 months prior to the index date in the follow-up period was important because abuse may result in greater use of medical resources prior to a formal diagnosis. Claims and costs 7 to 12 months before the incident diagnosis (the baseline period) were used in propensity score matching. A graphical depiction of the study period can be found in the eAppendix (eAppendices available at www.ajmc.com).
Two nonoverlapping cohorts of incident opioid abusers and nonabusers were selected. A 10% random sample of nonabusers was used. Patients in both cohorts had to be aged 18 to 64 years throughout the study period. In addition, we required continuous eligibility in non-HMO plans throughout the baseline and preindex follow-up periods (ie, up to and including the index date). After that, we followed patients for an additional 6 months, regardless of eligibility, to avoid bias from dropping patients who either died or lost healthcare coverage due to their abuse.
Propensity score matching of abusers and nonabusers was used to account for observable confounding factors associated with healthcare costs. A complete list of baseline characteristics, which were used as covariates in the propensity score model, and clinical codes can be found in the eAppendix. Abusers were matched 1:1 to nonabusers based on propensity scores ± one-fourth of a standard deviation using a greedy-matching algorithm.8 Baseline period characteristics for the postmatch cohorts were then compared to assess the quality of the match.
We compared abusers’ and nonabusers’ average healthcare costs to payers during 6-month increments of the study period: 1) the baseline period (7 to 12 months before the index event), the preindex period (1 to 6 months before the index event), and the postindex period (0 to 6 months following the index event). Medical costs were categorized by place of service: inpatient, emergency department (ED), outpatient/other (eg, skilled nursing facilities), and rehabilitation facilities (costs attributed to rehabilitation facilities include those for care received in inpatient and outpatient substance abuse facilities, including all treatment received for other, both abuse and non—substance abuse-related conditions). All costs were inflated to 2015 US dollars using the medical care component of the Consumer Price Index. To assess cost drivers, the top diagnoses contributing to excess medical costs were reported by 3-digit ICD-9-CM groupings.
We also calculated annual prevalence and incidence rates of opioid abuse for 2010 to 2014. Prevalence and incidence rates for 2015 were not available due to the incomplete year of data available at the time of the study. Prevalence rates reflect all patients with at least 1 claim for opioid abuse within a given year, whereas incidence rates reflect patients with incident diagnoses within a given year. The denominator for both incidence and prevalence comprises patients with at least 1 month of coverage within a year.
Incidence rates for those aged 18 to 64 years (ie, a rate of episodes per member) for 2012 to 2014 were averaged and combined with the estimate of annual, incremental excess costs due to opioid abuse to arrive at a PMPM estimate. Because annual excess costs of abuse were calculated for patients experiencing incident abuse episodes, incidence rates were considered more appropriate than prevalence for calculating a PMPM estimate of the burden of opioid abuse to payers.
Prematch Cohort Characteristics
Figure 1 describes the sample selection criteria that resulted in an analytic sample of 79,593 abusers and 1,817,395 nonabusers, for which prematch baseline characteristics were computed. The eAppendix Table (section i) compares baseline characteristics between the abusers and nonabusers prior to matching. Abusers were statistically different from controls on key demographic measures, with abusers more likely to be male (53.9% vs 45.5%; P <.0001) and younger (39.0 vs 43.0; P <.0001). There were also significant differences in the prematch comorbidity profile and healthcare resource utilization. During the baseline period, abusers were substantially more likely to have been diagnosed with abuse of nonopioid substances (15.3% vs 2.0%; P <.0001), major depressive disorder (8.5% vs 1.8%; P <.0001), other psychoses (7.9% vs 1.1%; P <.0001), and other mental health disorders (30.2% vs 9.1%; P <.0001). A complete list of the clinical codes and diagnoses used to construct these categories can be found in the eAppendix.
Overall, abusers used significantly more healthcare resources during the baseline period. For example, abusers spent more days in an inpatient hospital (0.8 vs 0.1 days; P <.0001) and more time in the ED (0.6 vs 0.1 visits; P <.0001) than nonabusers. These differences resulted in average baseline healthcare costs among abusers being nearly 4 times that of nonabusers ($8563 vs $2371; P <.0001).
Postmatch Cohort Characteristics
The propensity score procedure successfully matched 73,714 abusers 1:1 with nonabusers for a 93% match rate. These matched samples were well balanced with respect to the baseline characteristics (eAppendix Table [section ii]). There were 3 notable differences between the 2 cohorts, although the magnitude of these differences was small: 1) abusers were less likely to be the child of a primary beneficiary than nonabusers (27.3% vs 30.2%; P <.0001), 2) a larger proportion of abusers had extended-release opioid use during the baseline period (9.6% vs 7.9%; P <.0001), and 3) abusers had 2% higher costs than nonabusers during the baseline period ($7306 vs $7149; P = .0209).
Costs of Abuse
Figure 2 displays the average monthly excess costs during the study period. As observed in the original analysis, monthly excess costs increased during the 9 months leading up to the index date, from $7 eight months before to $774 in the month prior to the index date. Excess costs reached their maximum, $3628, in the index month, and although they dropped subsequently, remained well above baseline levels (between $1289 and $861) thereafter.
Tables 1 and 2 present the costs of abusers and controls during the two 6-month segments of the follow-up period (preindex, and postindex) by place of service, as well as the leading conditions driving excess medical costs. The average per-patient healthcare costs of abusers were $3018 more than nonabusers in the 6 months before the abuse episode and $7971 more than nonabusers in the 6 months after the abuse episode. Stated differently, abusers’ annual costs were estimated to be $25,069 compared with $14,080 for matched controls, resulting in an extra $10,989. In the 6-month preindex period, most excess healthcare costs were incurred in an inpatient setting (49%; $1471), followed by ED (18%; $531), rehabilitation facilities (17%; $503), outpatient (12%; $371), and prescription drugs (5%; $143). During the preindex period, the most costly conditions associated with opioid abuse included those related to nonopioid drug/alcohol abuse and dependence (19% of preindex excess costs), back pain (11%), and mental health-related disorders (6%).
In the 6 months after the index date, 53% ($4215) of excess healthcare costs could be attributed to care in a rehabilitation facility setting, 31% ($2429) in an inpatient setting, 8% ($610) in an ED setting, 5% ($432) on prescription drugs, and 4% ($284) on outpatient care. Similarly, opioid dependence, abuse, and poisoning (30%) and nonopioid drug/alcohol abuse/dependence (28%) were prominent excess cost drivers following the formal abuse diagnosis. As in the pre-index period, we also observed mental health-related (7%) and back pain-related (2%) conditions associated with the excess costs of abuse.
Figure 3 shows the diagnosed prevalence and incidence for opioid abuse, dependence, and overdose/poisoning combined. Prevalence rates between 2010 and 2014 nearly doubled (2.06 per 1000 in 2010 to 4.00 per 1000 in 2014), whereas incidence rates exhibited a flatter trend (1.46 per 1000 in 2010, to 2.46 per 1000 in 2014). As described in the Outcomes subsection of the Methods section above, combining incidence rates (the average incidence rate per 1000 for 2012 to 2014 is 2.16) and average annual excess costs ($10,989), we arrived at an estimate of $1.98 PMPM in costs to payers of opioid abuse.
To replicate the original analysis, the analyses above first examined the excess costs of opioid abuse and highlighted the substantial burden imposed on commercial payers: on average, a diagnosed opioid abuser had excess annual healthcare costs of $10,989. This figure is within the range of estimates from the existing literature3-5 and translates into $1.98 PMPM. This is comparable to the $14,810 in excess costs associated with opioid abuse in the original analysis.7 The difference between $10,989 and $14,810 may be explained by the unmatched patients in the Truven analysis. The match rate in Truven was 93% compared with 99% in Optum. The 7% of unmatched patients in Truven tended to be considerably more costly than those who matched; the much higher Optum match rate translated into including many of these costlier patients in that sample, resulting in higher excess cost estimates. Should the costlier Truven patients have matched, we may have observed very similar excess costs as those from Optum.
Next, the trajectory of excess costs by month was compared with that in the original analysis and confirmed the following: abusers’ mean healthcare costs began to build up prior to the formal diagnosis, spiked during the diagnosis month, and then flattened out. Notably, excess costs did not return to baseline levels. This pattern is similar to one recently observed among patients with cardiovascular disease9: for both low- and high-risk cohorts, incremental costs were observed accumulating in the year prior to a new cardiovascular event (2012: €148-€589; for comparison to our estimates, 2015: US$197-US$784 [2012 currency conversion rate: 1 US dollar = 0.81 euros10; the Bureau of Labor Statistics medical care component of the Consumer Price index was used to adjust to 2015 US$11]), peaked in the year of the new event (2012: €8346- €8663; 2015: US$11,108-$11,530), and persisted above pre-event levels for 2 years thereafter (2012: €1228-€1732 in the second year after the new cardiovascular event; 2015: US$1634-US$2305). As noted within the literature, such prolonged costs post index may be associated with considerable long-term costs. The assessment of longer-term incremental costs of opioid abuse may be a fruitful area of future research from which payers could benefit.
The final piece of the replication was to assess the drivers of excess costs. We found that a large proportion of excess costs was associated with opioid and nonopioid substance abuse (including alcohol), painful conditions (eg, intervertebral disc disorders, spondylosis, allied disorders), and mental health disorders (eg, episodic mood disorders, depression, anxiety). The associations between opioid and nonopioid substance abuse and between mental health disorders and substance abuse12,13 are consistent with the existing literature.
Other identified cost drivers were associated with vague diagnoses, including observation and evaluation for suspected conditions not found, general symptoms, and encounter for other and unspecified procedures and aftercare. One area for further research may be an evaluation of the extent to which such ambiguous diagnoses are associated with drug-seeking behavior. These predictors may help to identify improved treatment strategies for these complex patients by providing a rationale for an assessment for opioid use disorder.
Although it is well documented in the literature that the excess costs of opioid abuse impose a substantial burden on payers, this analysis contributes by highlighting the drivers of those costs. As evidenced in this analysis, patients given an opioid abuse diagnosis often present with numerous other complex and often costly conditions. This multifaceted clinical paradigm, along with recent increases in both the prevalence and incidence of opioid abuse, underscore the importance to public health and the healthcare system of diagnosing and treating opioid abuse and the multiple comorbidities that may accompany it. A more thorough understanding of the drivers of these costs may enable payers and policy makers to implement policies and patient care guidelines to more rapidly identify abuse and associated comorbidities, which may, in turn, help to lower costs.
First, the results rely on the accuracy of the administrative data. Therefore, any miscoding could affect our results, although we have no reason to suspect that any inaccuracies in the data affected the abusers or nonabuser control patients differently. Second, by definition, undiagnosed opioid abusers do not receive any of the ICD-9-CM diagnosis codes for abuse, and it is not out of the question that undiagnosed abusers may be included in the nonabuser cohort. Although the extent to which this applies to this particular sample is unknown, if undiagnosed abusers are more costly than a true nonabuser population, this would suggest that the estimated excess costs of diagnosed abuse understate the true excess cost differential between abusers and true nonabusers. Third, while our definition of abusers includes patients with overdose/poisoning diagnoses, the administrative data do not differentiate between patients who intentionally versus unintentionally overdose. Some unintentional overdoses may not reflect an abuse or misuse issue. Lastly, our findings may not generalize to noncommercially insured populations, although existing studies evaluating the excess costs of opioid abuse on other populations have generated similar estimates.14
This study confirms the findings contained in a recent publication: opioid abuse is costly to payers. Within a commercially insured population, opioid abuse, dependence, and overdose/poisoning were associated with $1.98 PMPM, or $10,989, in excess costs in the year centered around the initial diagnosis. The trajectory of opioid abuse—related costs is also robust across analyses: excess costs begin increasing 9 months prior to the index date, driven by nonopioid drug and alcohol abuse. Following diagnosis, costs were largely driven by the treatment of opioid abuse and nonopioid drug and alcohol abuse, as well as back pain and mental health-related conditions. Opioid abuse often occurs amidst a background of multiple comorbidities, including polysubstance abuse and other psychiatric disorders. Understanding the context in which opioid abuse occurs may promote a more comprehensive treatment approach among payers and providers.
The authors gratefully acknowledge the contributions of Caroline J. Enloe, Aliya Dincer, and Jessica Hanway.
Author Affiliations: Analysis Group, Inc (LMS, NYK, MLZ, ZBJ, HGB) Boston, MA; Purdue Pharma, L.P. (JCH) Stamford, CT.
Source of Funding: This study was funded by Purdue Pharma L.P.
Author Disclosures: Dr Howard was an employee of Purdue Pharma L.P. at the time of this study, and Drs Scarpati, Kirson, and Birnbaum, and Ms Zichlin and Mr Jia are employed by Analysis Group, Inc, which received research funding from Purdue Pharma L.P. for this study.
Authorship Information: Concept and design (LMS, NYK, ZBJ, HGB, JCH, MLZ); acquisition of data (NYK); analysis and interpretation of data (LMS, NYK, ZBJ, HGB, JCH, MLZ); drafting of the manuscript (LMS, NYK, JCH); critical revision of the manuscript for important intellectual content (LMS, NYK, HGB, JCH); statistical analysis (ZBJ, MLZ); and supervision (NYK, HGB).
Address Correspondence to: Lauren M. Scarpati, PhD, Analysis Group, Inc, 111 Huntington Ave, 14th Floor, Boston, MA 02199. E-mail: firstname.lastname@example.org.
1. Chou R, Fanciullo GJ, Fine PG, et al; American Pain Society-American Academy of Pain Medicine Opioids Guidelines Panel. Clinical guidelines for the use of chronic opioid therapy in chronic noncancer pain. J Pain. 2009;10(2):113-130. doi: 10.1016/j.jpain.2008.10.008.
2. Birnbaum HG, White AG, Schiller M, Waldman T, Cleveland JM, Roland CL. Societal costs of prescription opioid abuse, dependence, and misuse in the United States. Pain Med. 2011;12(4):657-667. doi: 10.1111/j.1526-4637.2011.01075.x.
3. Oderda GM, Lake J, Rüdell K, Roland CL, Masters ET. Economic burden of prescription opioid misuse and abuse: a systematic review. J Pain Palliat Care Pharmacother. 2015;29(4):388-400. doi: 10.3109/15360288.2015.1101641.
4. Meyer R, Patel AM, Rattana SK, Quock TP, Mody SH. Prescription opioid abuse: a literature review of the clinical and economic burden in the United States. Popul Health Manag. 2014;17(6):372-387. doi: 10.1089/pop.2013.0098.
5. Rice JB, Kirson NY, Shei A, et al. Estimating the costs of opioid abuse and dependence from an employer perspective: a retrospective analysis using administrative claims data. Appl Health Econ Health Policy. 2014;12(4):435-446. doi: 10.1007/s40258-014-0102-0.
6. Health research data for the real world: the MarketScan Databases. Truven Health website. http://truvenhealth.com/portals/0/assets/PH_11238_0612_TEMP_MarketScan_WP_FINAL.pdf. Published 2012. Accessed March 29, 2017.
7. Kirson NY, Scarpati LM, Enloe CJ, Dincer AP, Birnbaum HG, Mayne TJ. The economic burden of opioid abuse: updated findings. Manag Care Spec Pharm. 2017;23(4):427-445. doi: 10.18553/jmcp.2017.16265.
8. Austin PC. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011;46(3):399-424.
9. Hallberg S, Gandra SR, Fox KM, et al. Healthcare costs associated with cardiovascular events in patients with hyperlipidemia or prior cardiovascular events: estimates from Swedish population-based register data. Eur J Health Econ. 2016;17(5):591-601. doi: 10.1007/s10198-015-0702-0.
10. Yearly average currency exchange rates. Internal Revenue Service website. https://www.irs.gov/individuals/international-taxpayers/yearly-average-currency-exchange-rates. Updated December 12, 2016. Accessed December 22, 2016.
11. CPI databases. All urban consumers (current series) medical care component (CUUR0000SAM). Bureau of Labor Statistics website. https://www.bls.gov/cpi/data.htm. Published 2016. Accessed December 22, 2016.
12. Edlund MJ, Steffick D, Hudson T, Harris KM, Sullivan M. Risk factors for clinically recognized opioid abuse and dependence among veterans using opioids for chronic non-cancer pain. Pain. 2007;129(3):355-362.
13. Ross HE, Glaser FB, Germanson T. The prevalence of psychiatric disorders in patients with alcohol and other drug problems. Arch Gen Psychiatry. 1988;45(11):1023-1031.
14. Baser O, Xie L, Mardekian J, Schaaf D, Wang L, Joshi AV. Prevalence of diagnosed opioid abuse and its economic burden in the Veterans Health Administration. Pain Pract. 2014;14(5):437-445. doi: 10.1111/papr.12097.